CN101383151B - Digital audio quality reinforcing system and method - Google Patents

Digital audio quality reinforcing system and method Download PDF

Info

Publication number
CN101383151B
CN101383151B CN2007101498372A CN200710149837A CN101383151B CN 101383151 B CN101383151 B CN 101383151B CN 2007101498372 A CN2007101498372 A CN 2007101498372A CN 200710149837 A CN200710149837 A CN 200710149837A CN 101383151 B CN101383151 B CN 101383151B
Authority
CN
China
Prior art keywords
module
decoding
signal
gain
filtering
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN2007101498372A
Other languages
Chinese (zh)
Other versions
CN101383151A (en
Inventor
钟毅睿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
ZTE Corp
Original Assignee
ZTE Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by ZTE Corp filed Critical ZTE Corp
Priority to CN2007101498372A priority Critical patent/CN101383151B/en
Publication of CN101383151A publication Critical patent/CN101383151A/en
Application granted granted Critical
Publication of CN101383151B publication Critical patent/CN101383151B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Compression, Expansion, Code Conversion, And Decoders (AREA)

Abstract

The invention discloses a system and a method for enhancing digital audio quality. The system comprises a decoding gain calculation module, a post-filtering switch module, a longtime post-filtering module, a global gain control module and a pitch tracking module, wherein the decoding gain calculation module confirms the gain value of a received audio signal, and the post-filtering switch module judges whether the gain value confirmed in the decoding gain calculation module exceeds a threshold or not, if the gain value confirmed in the decoding gain calculation module exceeds the threshold, the longtime post-filtering module is started, or else a decoding audio signal is directly output; the longtime post-filtering module carries out the longtime post-filtering processing of the decoding audio signal and then output the decoding audio signal, the global gain control module carries out the whole gain control of the audio signal output by a longtime post filter, and the pitch tracking module derives a pitch period according to a decoding signal and pitch information and inputs the pitch period to the longtime post-filtering module as a control variable of filtering parameters. The invention decreases the complex degree of a digital audio post-processing process and can obtain favorable pitch enhancing effect.

Description

A kind of digital audio quality reinforcing system and method
Technical field
The present invention relates to the enhancing of the processing of digital audio and video signals, particularly audio quality.
Background technology
In common wideband digital audio equipment; after Network Transmission or audio coding decoding; the speech perception quality usually has obvious decline; so before sound signal is play; usually adopting needs to adopt certain disposal route that the audio frequency perceived quality is strengthened, for example the post-processing algorithm in audio codec.The aftertreatment of using in the AMR-WB codec of expansion is the optional fundamental tone enhancement post-processing of an a kind of frequency band algorithm at present.This method is carried out frequency division to decoded signal and is handled, and decoded signal is extracted radio-frequency component by high pass, fundamental tone is strengthened later decoded signal obtain low-frequency component by low-pass filter, finally merges two parts and obtains final output signal.Frequency division is handled the complexity that has increased demoder.Another is usually used in the fundamental tone technique for enhancing is to adopt comb filter, realizes that by comb filter fundamental tone strengthens.But prerequisite is the variation of the coefficient energy adaptive speech signal spectrum of comb filter, and this point implements difficulty.
Summary of the invention
The present invention is exactly in order to address the above problem, and has proposed a kind of succinct digital audio quality reinforcing system and method.Under the prerequisite that can obtain preferable DAB perceived quality, further reduced the complexity of Audio Processing process.
To achieve these goals, the technical solution used in the present invention is as follows:
A kind of digital audio quality reinforcing system comprises: the decoding gain calculation module, back filtered switch module, when growing after filtration module, global gain control module, fundamental tone tracking module; Wherein
Decoding gain calculation module: determine the gain of the sound signal that receives according to the signal energy that calculates of decoded audio signal of input; Filtration module after whether this yield value is sent into and be used for control in the filtered switch module of back and enable when long, an also controlled quentity controlled variable of postfilter parameter when long simultaneously;
Back filtered switch module: judge whether the gain of determining in the decoding gain calculation module surpasses predetermined threshold value, Filtering Processing after when if filtration module is grown after surpassing this threshold value and then enabling when long, otherwise directly export final sound signal, wherein, threshold value is set to 0.62.
Filtration module after when long: function is exactly the noise contribution of cutting down between fundamental tone harmonic wave peak value, strengthens the output audio quality; Export after the Filtering Processing after when this module is grown decoded audio signal; Filtered audio signal is input in the global gain control module and does gain control.
The global gain control module: the output audio signal of postfilter is exported final sound signal afterwards by total gain control that a global gain variable carries out during to length.
The fundamental tone tracking module: function is to try to achieve pitch period according to decoded signal and Pitch Information, be input to pitch period when long after in the filtration module as the controlled quentity controlled variable of filter parameter.
Further, the decoding gain calculation module is at first calculated the ratio of the signal energy of adjacent pitch period, then this ratio and the gain that decoding from code stream obtains is compared and get a wherein less yield value as final decoded signal.
Further, when long after filtration module, adopted full zero point after Filtering Model realize the back Filtering Processing.
Further, the fundamental tone tracking module, the global gain variable is tried to achieve later on by computing by the fundamental tone gain that the decoding gain calculation module calculates.
Further, the fundamental tone tracking module has adopted the fundamental tone tracking module in AMR-WB+ (AMR-WB of the expansion) codec.
A kind of digital audio quality Enhancement Method may further comprise the steps:
The first step: the yield value of determining the decoded signal that receives according to the signal energy that calculates of decoded audio signal of input;
Second step: whether judge the yield value of determining in the first step greater than predetermined threshold value, if greater than, then carried out for the 3rd step, otherwise carried out for the 5th step, wherein, threshold value is set to 0.62;
The 3rd step: Filtering Processing after when decoded audio signal is grown;
The 4th step: carry out being output as final signal after the global gain control;
The 5th step: the decoded audio signal of input is directly output as final signal.
Further, in the first step, calculate the ratio of the signal energy of adjacent pitch period, then this ratio and the gain that decoding from code stream obtains are compared and get a wherein less yield value as final decoded signal.
Further, in the 3rd step, adopted that Filtering Model realizes the back Filtering Processing after full zero point.
Compared with prior art, the invention provides a kind of digital audio quality reinforcing system and method, because the realization to the processing procedure of the configuration of the judgement of the threshold value in the filtered switch module of back and the coefficient in the post processing and filtering module is comparatively simple, reduce the complexity of processing procedure, and can obtain favorable pitch enhancing effect.
Description of drawings
Fig. 1 is the module relation diagram of each module of system of the present invention;
Fig. 2 is the process flow diagram of the embodiment of the invention.
Embodiment
Following accompanying drawing as a result is described in further detail the enforcement of digital audio quality reinforcing system proposed by the invention and method.
As shown in Figure 1, be the module relation diagram of each module of system of the present invention.
The decoding gain calculation module:
Determine the gain of the sound signal that receives according to the signal energy that calculates of decoded audio signal of input.At first calculate the ratio of the signal energy of adjacent pitch period, then this ratio and the gain that decoding from code stream obtains are compared and get a wherein less yield value as final decoded signal.Filtration module after whether this yield value is sent into and be used for control in the filtered switch module of back and enable when long, an also controlled quentity controlled variable of postfilter parameter when long simultaneously.
Back filtered switch module:
Judge whether the gain of determining in the decoding gain calculation module surpasses predetermined threshold value, if Filtering Processing after when surpassing this threshold value and then growing, otherwise directly export final sound signal.
This module is for when using aftertreatment, be provided with a judgment threshold, it is switching value, promptly, just do the back filtering operation, with the later audio signal output of filtering afterwards when the gain of determining in the decoding gain calculation module during greater than switching value, otherwise filtration module after not enabling when long, the direct final sound signal of output, wherein, the gain of determining in the decoding gain calculation module be current period with the signal energy in last cycle than and decoding obtains from code stream gain in less one.
Filtration module after when long:
The function of postfilter is exactly the noise contribution of cutting down between fundamental tone harmonic wave peak value when long, strengthens the output audio quality.This module is exported after to Filtering Processing after decoded audio signal (being the fundamental tone composite signal that decoding end decoding obtains) is when growing.The pitch period that the postfilter parameter calculates by the fundamental tone tracking module when long is done adaptive adjustment.Filtered audio signal is input in the global gain control module and does gain control.
Filtering Model realized the weakening of the noise between the fundamental tone harmonic wave is handled after this algorithm had only adopted full zero point.For the crest that guarantees the fundamental tone harmonic wave still at above frequency place, should be added on the corresponding frequency of the wave trough position place between those and fundamental tone harmonic wave zero point.By doing adaptive gain control, the perceived quality of final signal can improve greatly.
The global gain control module:
This module total gain control that the output audio signal of postfilter is undertaken by a global gain variable when long.The global gain variable is tried to achieve later on by computing by the fundamental tone gain that the decoding gain calculation module calculates.What the global gain control module was taked is adaptive gain control method, can ask for global gain in real time.The excessive audio frequency perceived quality of having introduced in the filtration module after when having avoided owing to length by the later sound signal of global gain control module that causes of gain fluctuation descends.
The fundamental tone tracking module:
Because Filter Design and filter effect are all closely related with the accurate prediction of pitch period, so this module has adopted the fundamental tone tracking module in the AMR-WB+ codec.Its function is to try to achieve pitch period according to decoded signal and Pitch Information, be input to pitch period when long after in the filtration module as the controlled quentity controlled variable of filter parameter.
As shown in Figure 2, it is the process flow diagram of the embodiment of the invention, among the figure, at first decoded audio signal is carried out gain calculating, the threshold value that will gain then and set in advance compares, if greater than threshold value, when just growing after filtering and global gain control and treatment, export final sound signal afterwards.If be not more than threshold value, just directly export final sound signal.
Specific implementation method is as follows:
The 1st step. determine the gain of the decoded signal that receives according to decoded signal.Specifically comprise:
Determine the ratio of the signal amplitude of adjacent pitch period, for AMR-WB+ in consistent, the computing method of amplitude ratio are
R = 0.01 + Σ i = 0 T s i 2 ( i ) 0.01 + Σ i = 0 T s i 2 ( i + T ) - - - ( 1 )
This ratio R and the gain that decoding from code stream obtains are compared and get a wherein less yield value g as final decoded signal aS wherein iThe decoded audio signal of this system of expression input, the pitch period value that T obtains for the fundamental tone tracking module.
The 2nd step. judge that whether the gain of determining in 1 surpasses predetermined threshold value, if surpass, then carries out for the 3rd step, otherwise carries out for the 4th step;
This algorithm for when using aftertreatment is provided with one and judges thresholding E Thr, promptly the signal energy when current period and last cycle compares E ComGreater than E ThrThe time, just do the back filtering operation, otherwise do not carry out this aftertreatment, directly export final sound signal.
Wherein, according to experiment, E ThrCan select 0.62;
The 3rd step. to exporting after the Filtering Processing after decoded audio signal (being the fundamental tone composite signal that the decoding end decoding obtains) is when growing;
Filtering Processing is exactly the noise contribution of cutting down between fundamental tone harmonic wave peak value after when long.In the codec based on CELP (Code Excited Linear Prediction), pitch predictor is just comprising the various information of pitch period, and the transition function of pitch predictor is H p(z)=1-g pz -TDescribe for convenient, it is rewritten as
1 B ( z ) = 1 1 - g p z - T - - - ( 2 )
G wherein pBe the fundamental tone gain, T is a pitch period.This fundamental tone composite filter is g at radius p 1/TCircumference on T limit arranged, and be spacedly distributed between each limit.Consider fundamental tone gain always on the occasion of, be 0,2 π/T so these limits lay respectively at phase angle ..., (T-1) * 2 π/T place, the peak value place of these positions and fundamental tone harmonic wave is one to one.
Filtering Model realized the weakening of the noise between the fundamental tone harmonic wave is handled after this algorithm had only adopted full zero point.For the crest that guarantees the fundamental tone harmonic wave still at above frequency place, should be added on the corresponding frequency of the wave trough position place between those and fundamental tone harmonic wave zero point, promptly at π/T ..., (2T-1) * π/T place is so the form of wave filter is
H(z)＝G 1(1+λz -T) (3)
Wherein, G 1Be the global gain control to this wave filter, λ is the local factor of adjusting.
The span of λ is generally between 0～1, and its value weighting degree between the signal of a pitch period that determined to be separated by after experiment, is got λ=0.1 in this algorithm.
The fundamental tone post-processing filter also can make the peak value at harmonic wave place that slightly variation is arranged between reduced harmonic in the noise.So, can be consistent with the real-time characteristic of voice spectrum for making filtering operation, should between reducing, find an equilibrium point by back caused signal skew of filtering operation and noise.For global gain G 1, what this module was taked is adaptive gain control method.
If k constantly this post-processing filter be input as x (n), be output as y (n), then the transition function from (3) can get
y(n)＝G 1×[x(n)+λ×x(n-T)] (4)
For unvoiced frame, according to strong periodicity as can be known, the waveform in its adjacent periods only is the difference on the amplitude, so can make
x(n-T)≈g a×x(n) (5)
G wherein aYield value for gain calculation module output.(5) are updated in (4), can get
y(n)≈G 1×[1+λ×g a]×x(n) (6)
By above derivation as can be seen, if do not do adaptive gain control, then wave filter can make output y (n) go out greatly much than input x (n) when finishing the fundamental tone enhancement post-processing of noise between reduced harmonic, and this can make the perceived quality of final signal descend greatly.So, global gain G 1Adaptation value can be set to
G 1 ≈ 1 1 + λ × g a - - - ( 7 )
In addition, because Filter Design and filter effect are all closely related with the accurate prediction of pitch period, so this module still adopts the fundamental tone tracking module among the AMR-WB+.
The 4th step. the fundamental tone composite signal that decoding end obtains is exported.
In the 2nd step and the 3rd step, establishing decoded audio signal is s i, the output signal of carrying out after the fundamental tone enhancement post-processing is y Out, then
y out = s i , if E com < E thr s i &CircleTimes; h , if E com &GreaterEqual; E thr - - - ( 18 )
The impulse response function of postfilter H (z) when wherein, h is for length.And this formula represents that the final sound signal of output has two kinds in the 4th step:
(1) a kind of is through after the 3rd back Filtering Processing that goes on foot, and the composite signal of process adaptive gain adjustment;
(2) another kind of for handling the directly sound signal of output without the 3rd step.

Claims (8)

1. a digital audio quality reinforcing system is characterized in that, comprising: the decoding gain calculation module, back filtered switch module, when growing after filtration module, global gain control module, fundamental tone tracking module; Wherein
Decoding gain calculation module: the yield value of determining the sound signal that receives according to the decoded audio signal of input; Filtration module after whether this yield value is sent into and be used for control in the filtered switch module of back and enable when long, an also controlled quentity controlled variable of postfilter parameter when long simultaneously;
Back filtered switch module: judge whether the gain of determining in the decoding gain calculation module surpasses predetermined threshold value, Filtering Processing after when if filtration module is grown after surpassing this threshold value and then enabling when long, otherwise directly export final sound signal, wherein, described threshold value is set to 0.62;
Filtration module after when long: output to after the Filtering Processing in the global gain control module after when decoded audio signal is grown and do gain control;
The global gain control module: the output audio signal of postfilter is exported final sound signal by total gain control that a global gain variable carries out during to length;
The fundamental tone tracking module: function is to try to achieve pitch period according to the decoded audio signal of input, be input to pitch period when long after in the filtration module as the controlled quentity controlled variable of filter parameter.
2. digital audio quality reinforcing system as claimed in claim 1, it is characterized in that, described decoding gain calculation module, at first calculate the ratio of the signal energy of adjacent pitch period, then this ratio and the gain that decoding from code stream obtains are compared and get a wherein less yield value as final decoded signal.
3. digital audio quality reinforcing system as claimed in claim 1 is characterized in that, described when long after filtration module, adopted full zero point after Filtering Model realize the back Filtering Processing.
4. digital audio quality reinforcing system as claimed in claim 1 is characterized in that, described global gain control module, and the global gain variable is tried to achieve later on by computing by the gain of the sound signal that the decoding gain calculation module calculates.
5. digital audio quality reinforcing system as claimed in claim 1 is characterized in that, described fundamental tone tracking module adopts the fundamental tone tracking module in the AMR-WB+ codec.
6. digital audio quality Enhancement Method may further comprise the steps:
The first step: the yield value of determining the sound signal that receives according to the signal energy that calculates of decoded audio signal of input;
Second step: whether judge the yield value of determining in the first step greater than predetermined threshold value, if greater than, then carried out for the 3rd step, otherwise carried out for the 5th step, wherein, described threshold value is set to 0.62;
The 3rd step: Filtering Processing after when decoded audio signal is grown;
The 4th step: carry out being output as final signal after the global gain control;
The 5th step: the decoded audio signal of input is directly output as final signal.
7. digital audio quality Enhancement Method according to claim 6, it is characterized in that, in the described first step, calculate the ratio of the signal energy of adjacent pitch period, then this ratio and the gain that decoding from code stream obtains are compared and get a wherein less yield value as final sound signal.
8. digital audio quality Enhancement Method according to claim 6 is characterized in that, in described the 3rd step, has adopted that Filtering Model realizes the back Filtering Processing after full zero point.
CN2007101498372A 2007-09-06 2007-09-06 Digital audio quality reinforcing system and method Expired - Fee Related CN101383151B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2007101498372A CN101383151B (en) 2007-09-06 2007-09-06 Digital audio quality reinforcing system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2007101498372A CN101383151B (en) 2007-09-06 2007-09-06 Digital audio quality reinforcing system and method

Publications (2)

Publication Number Publication Date
CN101383151A CN101383151A (en) 2009-03-11
CN101383151B true CN101383151B (en) 2011-07-13

Family

ID=40462953

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2007101498372A Expired - Fee Related CN101383151B (en) 2007-09-06 2007-09-06 Digital audio quality reinforcing system and method

Country Status (1)

Country Link
CN (1) CN101383151B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA2976485C (en) * 2010-07-02 2018-07-24 Dolby International Ab Audio decoder
CN103325380B (en) 2012-03-23 2017-09-12 杜比实验室特许公司 Gain for signal enhancing is post-processed
US8787403B2 (en) * 2012-05-14 2014-07-22 Texas Instruments Incorporated Audio convergence control facilitating bitpool value converging to stable level
CN110800050B (en) * 2017-06-27 2023-07-18 美商楼氏电子有限公司 Post linearization system and method using tracking signal

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101303858A (en) * 2007-05-11 2008-11-12 华为技术有限公司 Method and apparatus for implementing fundamental tone enhancement post-treatment

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101303858A (en) * 2007-05-11 2008-11-12 华为技术有限公司 Method and apparatus for implementing fundamental tone enhancement post-treatment

Also Published As

Publication number Publication date
CN101383151A (en) 2009-03-11

Similar Documents

Publication Publication Date Title
CN102169692B (en) Signal processing method and device
JP3653826B2 (en) Speech decoding method and apparatus
US8612218B2 (en) Method for error concealment in the transmission of speech data with errors
CN100550712C (en) A kind of signal processing method and processing unit
DE69934320T2 (en) LANGUAGE CODIER AND CODE BOOK SEARCH PROCEDURE
EP1775717B1 (en) Speech decoding apparatus and compensation frame generation method
US5873059A (en) Method and apparatus for decoding and changing the pitch of an encoded speech signal
AU2002221389B2 (en) Indexing pulse positions and signs in algebraic codebooks for coding of wideband signals
CN101383151B (en) Digital audio quality reinforcing system and method
CN102985969B (en) Coding device, decoding device, and methods thereof
FI95086C (en) Method for efficient coding of a speech signal
CN1113333C (en) Estimation of excitation parameters
JPH08328591A (en) Method for adaptation of noise masking level to synthetic analytical voice coder using short-term perception weightingfilter
CN103915100B (en) A kind of coding mode switching method and apparatus, decoding mode switching method and apparatus
JPH07248794A (en) Method for processing voice signal
KR100526829B1 (en) Speech decoding method and apparatus Speech decoding method and apparatus
DE19722705A1 (en) Method of determining volume of input speech signal for speech encoding
EP0843302A2 (en) Voice coder using sinusoidal analysis and pitch control
JP3070955B2 (en) Method of generating a spectral noise weighting filter for use in a speech coder
CN105593933A (en) Gain shape estimation for improved tracking of high-band temporal characteristics
CN101303858B (en) Method and apparatus for implementing fundamental tone enhancement post-treatment
CN101370056B (en) Digital audio automatic gain control method and its system
CN101533639B (en) Voice signal processing method and device
JP4343302B2 (en) Pitch emphasis method and apparatus
KR100421816B1 (en) A voice decoding method and a portable terminal device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20110713

Termination date: 20150906

EXPY Termination of patent right or utility model